INDEX

Explanations

phrases indicating a method or approach to doing something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating methods or ways to achieve something

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

avorite

-0.56

ono

-0.53

oppable

-0.52

inyl

-0.51

ibrary

-0.51

ilings

-0.51

 streng

-0.50

irie

-0.50

yss

-0.50

aples

-0.49

POSITIVE LOGITS

fare

1.05

ward

0.95

finding

0.92

 forward

0.91

station

0.91

forward

0.90

of

0.87

to

0.85

point

0.84

points

0.81

Activations Density 0.034%

phrases indicating a method or approach to doing something

phrases indicating methods or ways to achieve something

No Comments

No Known Activations